Well the part of the article that quotes the actual job listing:

Part of that role, compensated at between $180,000 and $230,000, will be to “identify, assess, track, and investigate global threats including geopolitical instability, terrorism, crime, activism, nation-state targeting of the AI sector, and emerging security trends, including deep-dive research and OSINT collection on specific threats, actors, and events” (emphasis added).

But yeah, not a good look especially if you consider who decides what "terrorism" and "crime" mean in the US right now.

And now for a tangent I wrote that you can just skip if you're not interested:

The part about specifically tracking AI activists was that they watch for anti-AI protests anywhere where their executives go so they know to stay out of the way, but... How do they not realize that they've brought this upon themselves? Market AI as "doing everyone's job" and "super dangerous" and of course people will be pissed at you. Dario himself is the one constantly parroting how the next Claude model is so dangerous it needs to be kept in adamantium chains and whatnot. Like your industry is already pissing people off with the data centers and the cost of computer parts, why you gotta tell everyone you're working to put us all out of jobs?

Jfc I'm a software engineer with no marketing or PR experience and I think even I could do better PR than all these companies supposedly worth billions. If OpenAI is already farming outrage, just position yourself as the more responsible alternative and you'll still get sales without your execs needing to fear for your lives... Occasionally release a smaller, less capable version of Claude as open weight to win mindshare among developers, the people actually burning tons of tokens right now. Like Alibaba does with Qwen, just don't release the full size model like Alibaba does if you're afraid people are going to run it in competition with you.